← Back to blog
President Trump With Tech Leaders as Lincoln Looks Down

President Trump With Tech Leaders as Lincoln Looks Down

Rhetoric-reality inversion

September 5, 2025


Designing a system that will last is unimaginably hard. Designing a system which lasts and whose mission and actions continue to align is impossible. I've had a theory of vision entropy for a while, where the mission of a system degrades over time as power changes hands. But recently, I've seen a pattern where the mission of systems tend to decay to a point where they represent the antithesis of what they were designed to represent. That is definitely the most entertaining outcome.

I've been running an experiment for a few weeks now. I bring it up when I don't know what to say in a conversation (often). I ask the other person, "what form of government do we live in on paper versus in practice?" A very normal first-date type question.

Most people who've taken high school US History say "we're in a democracy/republic" but are unsure of what we are functionally. Everyone in my aristocratic bubble is aware of how campaign finance and redistricting have fucked up our mirage of a democracy but they still call the US a democracy. Aristotle would call our constitutional republic an elected aristocracy, but that disturbs most people. I always tell them that we're in an oligarchy functionally, but I never get much valid pushback—it's an open secret—but everyone seems to be chill with it.

The force of vision entropy has the outcome of rhetoric-reality inversion. The metric to track this is the rhetoric-reality error.

I see this inversion, between rhetoric and reality, everywhere:

The best way to analyze a system isn't in its "rhetoric" or "reality" but in the error between the two. There are two ways to look at this:

  1. the error may seem hypocritical at a glance, but it may be necessary for the success of the system (democracy needs democratic myths to stabilize oligarchy). It's irrational to try to prevent the inevitable, so just cover it up with a myth.
  2. the error represents the failure of the current system to match its intended ideal, and at a certain degree of error, the system must be overthrown to bring it back to the ideal. This represents the march of history

Nations demonstrate emergent effects from their constituent individuals:

History is obviously in some sort of steady-state equilibrium between rhetoric and reality, between anarchy and autocracy. In machine learning, the machine only learns if the error decreases over time, which means the predictions get closer and closer to the reality. If there was a system where the error only got worse over time, the model would need to be retrained. In our society, the "retraining" (revolutionary) ethic seems to relegated to a small number of outcasts who lead insurrections while the vast majority of us are fine with it. I wonder what degree of error would warrant a Revolution or if we should just be fine with the myth.


Nearby